Chapter 90. Fundamentals of Cutaneous Photobiology and Photoimmunology

Knowledge of the interaction of sunlight with the skin is fundamental to understanding the pathogenesis, diagnosis, and treatment of more than 100 cutaneous disorders. Whenever ultraviolet (UV) or visible radiation is used to diagnose or treat a skin condition, important principles of photophysics involving absorption and emission of light underlie the success of the therapy. Sunscreen recommendations rely on an understanding of solar UV radiation and the ways in which the causative wavelengths can be minimized. Skin cancer is an epidemic clinical problem, whose pathophysiology necessitates comprehension of the photophysical, photochemical, and photobiologic events described in this chapter.

Almost every ancient civilization worshipped a god of the sun whose healing powers were believed to be broad reaching. Even today, sun exposure is widely felt to induce a sense of well-being. In addition, sunlight is important for the synthesis of vitamin D3 and the setting of internal clocks. On the negative side, sunlight causes deleterious acute and chronic inflammatory skin reactions, skin cancer, and photoaging, and can elicit adverse reactions to certain drugs (see Chapters 91, 92, 109, 112). Although the sun is a major source of the UV and visible radiation that interacts with human skin, UV and/or visible radiation are also emitted from common sources such as fluorescent lights, incandescent bulbs, photocopy machines, and phototherapy lamps. Tanning salons are another familiar example. Thus, UV and visible radiation are a constant part of the human environment and play a role in health, disease, and therapy. Photodermatology is the study of this interaction between human skin and UV and visible radiation. To understand the responses of skin to UV and visible radiation, it is essential to be acquainted with the principles governing the interaction of these wavebands with biomolecules in the skin.

When UV and visible light photons reach the skin surface the energy of the radiation is transformed into an observable response as shown schematically in Figure 90-1. First, the radiation must penetrate to the appropriate level in the skin where it is absorbed by molecules in the skin, termed chromophores. Photochemical reactions then convert the chromophores into new molecules, the photoproducts. These photoproduct molecules stimulate cellular signal transduction pathways leading to biochemical changes that culminate in cellular effects, such as the proliferation, secretion of cytokines, and apoptosis that are responsible for the observed acute skin responses.

UV radiation and visible light are portions of the electromagnetic (EM) spectrum, which includes a wide range of wavelengths, from high-energy X-rays to low-energy microwaves and radio waves (Table 90-1; Fig. 90-2). The UV waveband is of special interest because dozens of skin disorders are aggravated by these wavelengths and, many popular therapies, such as UVB phototherapy (see Chapter 237) and psoralen and ultraviolet A light (PUVA) photochemotherapy (see Chapter 238), use sources emitting UV radiation. Visible light encompasses those wavelengths perceived as color by the human eye and is also frequently used in therapies, such as blue-light amino levulinic acid–photodynamic therapy (PDT; see Chapter 238), and is emitted by several lasers intended to target cutaneous chromophores (see Chapter 239).

Ultraviolet Radiation

For medical photobiology, the UV range (200–400 nm) is subdivided into UVA, UVB, and UVC (see Table 90-1 and Fig. 90-2). A division was made at 290 nm because wavelengths from the sun shorter than 290 nm are absorbed by ozone in the stratosphere and do not reach the earth's surface at sea level. Wavelengths in the range of 200 to 290 nm are referred to as UVC or germicidal radiation. These wavelengths are strongly absorbed by DNA and therefore can be lethal to viable cells of the epidermis or to bacteria. UVC lamps emit at 254 nm and are used for air and water purification. Care must be taken to avoid exposure of eyes and skin to UVC radiation because of the danger of UV keratitis and mutation.

The range 290 to 320 nm is known as UVB and is often referred to as mid-UV or sunburn spectrum. It includes the biologically most active wavelengths reaching the earth's surface. The UVB constitutes only approximately 5% of the UV and 0.5% of total radiation reaching the earth's surface; the exact amount varies markedly with the time of day, season, cloud conditions, and other factors. Ordinary window glass blocks UVB. Most sunscreens efficiently reflect or absorb these wavelengths, and the sun protection factor (SPF) is primarily based on testing against this waveband. However, it is important to note that different wavelengths within these subdivisions can elicit greatly varying biologic responses. For example, consider the response of skin to two wavelengths in the UVB range, 297 and 313 nm. Radiation at 297 nm is nearly 100 times more erythemogenic than 313-nm radiation1 and more effectively causes DNA damage and photocarcinogenesis.2 An example relevant to phototherapy is the great efficacy of certain UVB wavelengths for treatment of full UVB spectrum.3,4 Narrowband UVB (311 nm) and excimer laser (308 nm) are used to treat psoriasis because these wavelengths are more effective than other portions of the UVB spectrum.3,4

Long-wave UV or UVA (320–400 nm) is sometimes referred to as black light because it is not visible to the human eye but causes certain substances to emit visible fluorescence. Approximately 95% of the UV radiation reaching the earth's surface is UVA. As described for UVB radiation, the response of skin to UVA is not constant across all wavelengths from 320 to 400 nm.3 In fact, UVA has been divided into UVA I (340–400 nm) and UVA II (320–340 nm) because the latter band is more damaging to unsensitized skin than the longer wavelengths. Although most sunscreens offer their greatest protection against UVB wavelengths, those products said to be “broad-spectrum” indeed have considerable ability to protect against UVA wavelengths, although no widely used formal measure equivalent to the SPF rating yet exists. Generally, the higher the SPF, for example, 45 or greater, the more likely one is going to find some UVA protection in that preparation. Newer sunscreen agents enhance protection against damaging UVA rays (see Chapter 223). Because of its predominance in the UV radiation reaching the earth's surface, protection against UVA is quite important for minimizing adverse cutaneous effects such as photoaging and carcinogenesis.

Visible Radiation

The visible spectrum (400–760 nm) is defined by the wavelengths that are perceived as color by the retina. Specific colors are associated with different wavelengths, as shown in Fig. 90-2. Skin responses to visible light generally require photosensitization. For example, in PDT, dyes absorbing long wavelength visible light (red) are used (see Chapter 92). Because they take advantage of the absorption spectra of endogenous chromophores, intense pulsed visible light sources such as lasers are used to treat vascular, pigmented and other lesions without application of a photosensitizing dye (see Chapter 239).

Other Wavebands of Electromagnetic Radiation

X-rays and γ rays occupy the short wavelength (high energy) end of the EM spectrum, and infrared radiation (IR) is found at longer wavelengths (lower energy) than visible radiation (see Table 90-1). X-ray and γ rays ionize molecules (remove electrons) indiscriminately and are known as ionizing radiation, the subject of radiobiology (see Chapter 240). In radiation therapy of tumors, these wavelengths kill tumor cells by ionizing water molecules and producing free radicals that damage DNA. IR has lower energy than visible light. It is subdivided into IR-A (760–1,440 nm), IR-B (1,440–3,000 nm), and IR-C (3,000 nm–1 mm). IR-A penetrates into the dermis and causes skin damage whereas IR-B and IR-C are felt as heat. Recent studies suggest that IR-A wavelengths might also be therapeutic.5

Certain principles are better illustrated by conceptualizing EM radiation as waves, whereas others are better understood by thinking of it as packets of energy called photons. These two descriptions are complementary.

When considering EM radiation as a wave, it is described as oscillating electric and magnetic fields at right angles to each other and to the direction of propagation. Consequently, it may be described either by its frequency (number of oscillations per second) or by its wavelength (distance traveled per oscillation). Frequency and wavelength have an inverse relationship, which is expressed as: ν = c/λ, where ν = frequency (number of oscillations per second), c = speed of light (3 × 108 m/s), and λ = wavelength in meters.

EM radiation also may be described as a stream of discrete packets of energy known as quanta or photons. The amount of energy in a photon (quantum) is directly proportional to the frequency of the radiation and inversely proportional to its wavelength, as expressed by Planck's law: E = hc/λ, where E = the energy of the photon in joules (J), h = Planck's constant (6.626 × 10−34 J/s), c = speed of light, and λ = wavelength in meters. This relationship shows that the energy of the photon increases when the wavelength is shorter and decreases when the wavelength is longer. For example, a 300-nm UVB photon has twice the energy of a 600-nm yellow photon.

Sunlight

The shortest wavelength of the solar spectrum reaching the earth's surface at sea level is approximately 290 nm, although slightly shorter wavelengths are detected at high altitudes. Depending on the geographic location and the season, it has been estimated that sunlight produces between 2 and 6 mW/cm2 of UV radiation between 290 and 400 nm. Filtering of wavelengths less than 290 nm by ozone is a very important process because the shorter UVC wavelengths are highly damaging to animals and plants. Because the transmission of solar UVC and UVB through the atmosphere varies exponentially with ozone concentration, small changes in the ozone layer may result in hazardous increases in UV irradiance at the earth's surface.6 For example, calculations of the effects of ozone depletion predicted a doubling of skin cancer incidence by 2,100 ad even if the Montreal Protocol restrictions on ozone-depleting substances were followed.7

The UV Index developed by the National Weather Service and the US Environmental Protection Agency represents an attempt to quantify the risks attendant with solar radiation at a given time and place. Among other factors, the effects of altitude, latitude, season, and clouds are considered. The scale goes up to 15, and any UV Index greater than 10 is considered a high-risk day for possible overexposure in that locale.

Artificial Sources of Ultraviolet and Visible Radiation

Skin is exposed to UV and visible radiation from a wide variety of sources in daily life and from a different set of light sources for therapy and diagnosis. Specific information should be obtained from the manufacturer before using a new light source.

Incandescent Sources

Incandescent light sources include conventional electric light bulbs, flood lamps, and some quartz iodide bulbs. In these lamps, an electric current passing through a metal filament heats the filament, causing it to emit mostly visible and IR. Only the occasional patient with solar urticaria, chronic actinic dermatitis, or some porphyrias is bothered by the output of ordinary incandescent sources. Tungsten-halogen incandescent lamps, often used as flood lamps, emit UVA and visible radiation. Some quartz iodide incandescent lamps produce significant UVA and some UVB emission.

Arc Sources

Arc sources include xenon lamps, medium- and high-pressure mercury (hot quartz) lamps, fluorescent lamps, and halide lamps. In arc lamps, electrons are driven through a gas by a potential difference between two electrodes. The gaseous molecules are ionized and subsequently release EM radiation. Radiation is emitted preferentially at specific wavelengths (emission lines) as well as in a continuum, that is, all wavelengths are emitted rather than just specific wavelengths. The wavelengths and relative power at each wavelength depend on the gas used, the arc temperature, the pressure within the lamp, and the lamp wall material.

Xenon arc lamps emit both UV and visible radiation and are now the most common sources used in solar simulators. Photoprovocation testing for polymorphic eruption is often done with such sources using filters to limit the wavelengths. Xenon arcs are also used in some phototherapy and photobiologic research applications. In a xenon lamp, xenon gas under 20–40 atmospheres (atm) of pressure produces intense visible and UV radiation. At these pressures, the xenon spectrum becomes a continuum.

The wavelengths emitted by mercury arc lamps are strongly influenced by the pressure of the gas within the envelope. Low-pressure mercury germicidal lamps emit 85% of the radiant energy at 254 nm. Because the operating temperature is low, they are also known as cold quartz lamps. With increasing pressure (1 atm), the primary 254-nm emission is absorbed by other mercury atoms within the lamp and reemitted at longer wavelengths (297, 302, 313, 334, and 365 nm, and visible wavelengths). With further pressure increases (2–100 atm), these spectral lines broaden and decrease relative to the intensity of the continuous spectral background. In medical practice, medium- and high-pressure mercury (hot quartz) lamps are generally used as sources of UVB, although their spectral power distribution is mainly in the UVA and visible range.

The commonly used UVB sunlamps (see Chapter 237) and UVA lamps for PUVA therapy (see Chapter 238) are fluorescent lamps. They are, in essence, modified low-pressure mercury arc lamps. The inner surface of the glass tube is coated with a phosphor, which absorbs the 254-nm radiation and reemits the energy at longer wavelengths. The chemical composition of the phosphor determines which wavelengths are reemitted. In general, fluorescent lamps have a relatively wide, bell-shaped emission spectra, with emission lines of mercury superimposed, and are referred to as broadband light sources.

Fluorescent sunlamps (type FS) emit mainly in the UVB range (Fig. 90-3). They are often referred to as UVB lamps, even though they emit a portion of UVA radiation, because the therapeutically significant radiation is in the UVB range (Table 90-2). A fluorescent lamp with a major emission peak at 311 nm (Philips TL01) was developed for use in phototherapy (see Fig. 90-3).8,9 This lamp is an efficient source for psoriasis phototherapy because, compared to a conventional UVB lamp, the energy emitted almost entirely overlaps with the action spectrum for clearance of psoriasis.3 Interestingly, this lamp has also been successfully used for the treatment of vitiligo, atopic dermatitis, and polymorphic light eruption, all of whose action spectra are unknown and may possibly differ from that for psoriasis.

High-intensity, UVA-emitting fluorescent lamps are most often used in PUVA therapy of psoriasis, vitiligo, and other skin diseases. The most efficient UVA source for PUVA therapy would maximize the emission of 320- to 360-nm radiation for photoactivation of psoralen molecules, while minimizing the UVB emission. Table 90-2 shows the percentage of total emission that is in the 320–360-nm range from several UVA-emitting light sources that are used in phototherapy units.

Wood's lamps are small, low-pressure, UVA-emitting fluorescent lamps with a UVA-transmitting, visible-absorbing glass envelope. They are useful in clinical practice because, after UVA absorption, the fluorescent emission from normal and abnormal components of skin, hair, teeth, and urine may be diagnostic (e.g., in porphyria, vitiligo, and fungal infection).

Halide lamps emit a high-intensity continuum in the UVB and especially the UVA range. With appropriate filters, this lamp is used increasingly as a UVB, as well as a UVA, source for phototherapy (in particular, UVA I therapy) and photochemotherapy. The metal halide lamp usually consists of a high-pressuremercury lamp with metal halides as additives. The continuous range of wavelengths emitted from halide lamps differentiates them from medium-pressure mercury arc lamps that emit in narrow wavelength ranges. Approximately 20% of the emission spectrum can be UVA radiation.

Lasers

(See Chapter 239.)

Lasers produce intense beams of monochromatic (single wavelength) radiation. The laser operates by exciting molecules to a metastable excited state from which photon emission is stimulated by a subsequent photon incident on the excited molecule. The emitted photon and the stimulating photon are then each capable of stimulating emission of yet other excited molecules, eventually producing an avalanche of photons of the same wavelength, phase, and direction of propagation. Different lasers emit UV, visible, or infrared wavelengths and may operate as either continuous or pulsed sources.

To treat patients with the appropriate amount and wavelengths of UV or visible radiation, it is important to become acquainted with how the radiation is quantified and related to the exposure time. The basic unit of EM energy is the joule (J). Power is the rate of energy flow, joules per second (J/s), usually expressed as watts (W). The rate at which the radiant energy is delivered to a surface, such as skin, is expressed as the power delivered per unit area of surface (power/area; W/cm2). This quantity is called irradiance. The total radiant energy delivered per unit area of skin surface over a period of time is called the exposure dose or fluence and is the product of irradiance and time:

Irradiance (w/cm2) × time (s) = Exposure dose or fluence (J/cm2)

For most responses to UV and visible radiation, it is the fluence at particular wavelengths that determines the magnitude of the response. The irradiance generally does not affect the response using conventional light sources.

The irradiance delivered by a source as a function of wavelength is called the spectral irradiance and is expressed as units of irradiance per nanometer [(W/cm2)/nm]. A spectroradiometer is used to measure the spectral irradiance of a light source. When measuring a light source's irradiance over a given spectral region, a detector weighted to the wavelengths of interest is most useful. For example, a broadband radiometric measurement of wavelengths less than 315 nm provides a rough indication of the erythemally effective wavelengths emitted by a source of UV radiation.

There is, however, no substitute for knowing the full spectral irradiance delivered by a source, as determined by a spectroradiometer. For example, in phototesting, only the wavelengths of interest should be used. To assess endogenous UVA sensitivity, the UVB portion of a source's emission, if present, must be filtered out so that the more erythemogenic UVB wavelengths do not lead to a falsely lower determination of erythema threshold in the UVA range.

When UV and visible radiation strike the skin, part is remitted (reflected and scattered), part is absorbed by chromophores in various layers, and part is transmitted inward to successive layers of cells until the energy of the incident beam has been dissipated (Figs. 90-4 and 90-5). The two major processes limiting the penetration of UV and visible radiation into skin, absorption and scattering, vary with wavelength.10

UV wavelengths less than 320 nm are readily absorbed by proteins, DNA, and other components of epidermal cells. Along with scattering, this absorption accounts for the low penetration of these wavelengths into skin (see Fig. 90-4). For example, approximately 10% of 300-nm radiation and 50% of 350-nm radiation reaches the dermal–epidermal junction in fair skin. Between 5% and 10% of incident light is reflected by the outer surface of the stratum corneum. This surface or so-called specular reflectance is relatively constant for all visible wavelengths and accounts for the surface appearance of skin, which is especially glossy if the surface is smooth, wet, or oily. In contrast, if the surface is irregular, the light is scattered and the skin appears dull or “rough.” Moisturizers applied to the skin reduce this rough appearance by smoothing out the many air–surface interface irregularities and making the skin look shinier.

Scattering includes any process that deflects the path of optical radiation. For example, skin with scales, as in psoriasis, scatters more light than does normal skin. During phototherapy, application of emollients to the psoriatic plaques helps reduce the scattering of UV radiation and allows more of the effective wavelengths to penetrate into the viable tissue.

Melanin, which absorbs relatively uniformly over the visible wavelengths and is normally present only in the epidermis, acts largely as a neutral density filter to diminish remittance from the dermis. The greater overall melanin content in darker skin absorbs more visible light and therefore causes the skin to appear darker as there is less light remitted back to the observer. Blood (hemoglobin) within the dermis absorbs the shorter (blue) visible wavelengths, diminishing these spectral regions, and giving a reddish hue to our perception of the total remittance. Abnormal location and quantity of these or other pigments account for the appearance of the skin in pathologic conditions (e.g., melasma with extra pigment in the epidermis and/or dermis, vitiligo with an absence of epidermal melanin).

When a photon is taken up by a chromophore, it is said that absorption has occurred. The specific wavelengths absorbed by each molecule (called an absorption spectrum) are characteristic of the structure of the molecule (i.e., the arrangement of the nuclei and electrons). Only radiation that is absorbed by the chromophores can initiate biologic responses.

The absorption spectrum of a compound is a plot of the probability of absorption of photons of a specific wavelength on the Y-axis against wavelength on the X-axis (Fig. 90-6). The wavelengths that have the highest probability of absorption are called absorption maxima, λmax, (e.g., DNA, λmax = 260 nm; porphyrins, λmax = 400–410 nm). Many of the biomolecules that absorb in the UVB spectrum actually have absorption maxima at shorter wavelengths in the UVC range (see Fig. 90-6). These include the purine and pyrimidine bases in DNA and RNA (λmax ∼260 nm) and 7-dehydrocholesterol (λmax = 285 nm). Many endogenous cutaneous chromophores absorb UVA or a combination of UVA and visible radiation, including hemoglobin (λmax = 410 nm) and bilirubin (λmax = 450 nm). Melanin absorbs throughout the UV, visible, and near IR spectra without a distinct absorption maximum wavelength. The absorption spectra of few chromophores extend into the IR-A waveband, an exception being cytochrome c oxidase, which is last enzyme in the mitochondrial respiratory electron transport chain.

After absorbing the energy of a photon, the chromophore is in an “excited state,” which exists for only a very short time before reacting with nearby molecules. The products of these reactions initiate signal transduction processes that lead to the observed responses in skin.

Excited State Molecules

Normally, molecules are in the so-called ground state and have a certain distribution of electrons in space around the nuclei of their atoms. For each molecule, a series of electronic states with higher energies and different distributions of electrons also exist; these are called excited states. When a molecule in the ground state absorbs the energy of a UV or visible photon, the molecule is promoted to an excited electronic state (Fig. 90-7). According to quantum mechanics, only certain energy gaps are allowed to exist between electronic states. Consequently, a molecule can absorb photons only with certain energies; this results in a unique absorption spectrum for each molecule.

A molecule exists in this first excited state formed for a very brief period. It is called a singlet excited state and exists for a few nanoseconds. The molecule may return to the ground state by emitting light (fluorescence) or releasing the energy as heat by a process called internal conversion (see Fig. 90-7). Alternatively, the singlet excited state may undergo a chemical reaction to form a photoproduct, or it may convert into another excited state with lower energy, the triplet excited state, by a process called intersystem crossing (see Fig. 90-7). Singlet and triplet excited states differ in the spins of a pair of electrons in an orbital. The ground state is almost always a singlet state. The triplet excited state may exist for a longer time (i.e., a few microseconds). It may emit light (phosphorescence), undergo a chemical reaction, or return to the singlet ground state by intersystem crossing (see Fig. 90-7).

These excited state processes are responsible for the effectiveness of light for diagnosis and therapy. For example, fluorescence occurs every time a Wood's light is used. UVA emitted from this lamp causes autofluorescence of dermal collagen fibers. To the examining physician, this fluorescence is viewed through the overlying epidermis. Thus, any epidermal lesions such as lentigines tend to have their borders accentuated by contrast because the fluorescence is observed most brightly around the lesion. The heat generated by internal conversion is responsible for the effects of pulsed lasers (see Chapter 239).

Photoproducts

During a photochemical reaction, the excited state molecule is transformed into a new, stable molecule called the photoproduct. Photoproducts in DNA are important for UVB-induced responses in skin. The photoproduct molecule may be produced entirely by rearrangement of the bonds in the chromophore, for example, the formation of previtamin D3 from 7-dehydrocholesterol or in DNA of a four-membered ring structure called a cyclobutyl pyrimidine dimer (CPD) from adjacent thymines or cytosines. These DNA photoproducts lead to mutations and development of nonmelanoma skin cancers (see Chapter 112). Alternatively, the chromophore may form covalent bonds with another molecule in the cell, for example, the formation of covalent adducts between psoralen and DNA. The phototherapy of diseases such as psoriasis makes use of this photochemical reaction (see Chapter 238). In addition, many skin responses to UV and visible light are initiated when the excited state chromophore transfers its energy to oxygen to form singlet oxygen or transfers an electron to form superoxide anion. These forms of oxygen react with cellular molecules, which often initiates intracellular signal transduction leading to the inflammation seen in sunburn and drug phototoxicity.

Photochemical reactions vary in efficiency. Not every chromophore molecule that absorbs a photon undergoes a photochemical reaction because the excited states may fluoresce or follow other pathways (see Fig. 90-7). The term quantum yield indicates the likelihood that one of these processes occurs. For example, the quantum yield for forming a certain photoproduct is:

Photosensitization

When certain drugs (e.g., tetracyclines, fluoroquinolones, psoralens) and dyes absorb UV and/or visible radiation, delayed erythema or inflammation is observed (see Chapter 92). This phenomenon is called photosensitization, and the dyes and drugs are called photosensitizers. Photosensitization requires the presence of oxygen in most cases and the initial photoproducts are reactive oxygen species (ROS). ROS are small molecules and free radicals including singlet oxygen, hydrogen peroxide, superoxide anion and hydroxyl radical. These ROS oxidize unsaturated lipids in membranes, certain amino acids in proteins (histidine, methionine, tryptophan, cysteine), and guanine in nucleic acids. The oxidized products initiate signal transduction processes leading to inflammatory mediators, for example, prostaglandin E2 (PGE2) and tumor necrosis factor (TNF)-α, that produce the inflammation observed as erythema.

A photosensitizer molecule in an excited triplet state can transfer its energy to an oxygen molecule, thereby generating the ROS called singlet oxygen (1O2). Singlet oxygen is relatively long-lived for a singlet state (<4 microseconds) and reacts with other cellular molecules to generate additional ROS. The skin photosensitization produced by PDT drugs and by protoporphyrin IX, the porphyrin that accumulates to abnormally high levels in red blood cells of patients with erythropoietic protoporphyria (EPP), involves initial formation of singlet oxygen. Oxidation of membrane lipids is believed to initiate the signal transduction processes, leading the wheal and flare response observed in EPP photosensitivity. Singlet oxygen is also involved in the phototoxicity mechanisms for some phototoxic drugs. However, the observed skin phototoxicity responses often differ from that seen in EPP photosensitivity, possibly because of the presence of the photosensitizing agents in different tissue locations (e.g., epidermis, dermis, or vasculature). The photosensitizers may also be in different locations in the cells (i.e., nucleus, mitochondria, or cell membranes), which would influence the photoproducts formed and the subsequent cutaneous effect observed. Endogenous UVA-absorbing chromophores generate singlet oxygen in keratinocytes and fibroblasts and in skin.11

Endogenous chromophores and some exogenous photosensitizers generate other ROS.12 For example, hydrogen peroxide and superoxide anion are produced after photosensitized and direct photon damage to mitochondria. The UVB or UVA radiation also activates ROS-producing enzymes such as nicotinamide adenine dinucleotide phosphate (NADPH) oxidase in keratinocytes and fibroblasts.13,14 The UVA-induced responses in normal skin and in photosensitivity conditions are generally dependent on production of ROS. Recent studies suggest that ROS are important in the development of many UVB-initiated responses. Antioxidants act by quenching (i.e., chemically reacting with and removing) ROS and other free radicals before they can damage cellular molecules.

Sunburn and tanning are the most obvious responses elicited by exposure of skin to a single dose of UV radiation. Many other responses are less visually apparent, but have important physiologic effects including formation of vitamin D3, altered immune responses, production of antimicrobial peptides and disruption of the epidermal barrier function. In addition, the cumulative effect of repetitive exposures gives rise to chronic skin changes including skin cancer development and photoaging. All of these responses result from a complex burst of UV-induced activity in the epidermis and dermis involving cytokines, neuropeptides, prostaglandins, ROS, and altered expression of at least 600 proteins.11,15–17

This section centers on three different parts: (1) UV-induced inflammation (sunburn), (2) vitamin D synthesis, and (3) altered immune responses. Two other acute responses to UV, (1) tanning (melanogenesis) and (2) abnormal acute responses to UV, are described in Chapters 72 and 91, respectively.

Acute Inflammation—Sunburn

Erythema is the most visually apparent indicator of UV-induced skin inflammation. Increased blood flow in the superficial and deep vascular plexus is responsible for the appearance of erythema and also may increase skin temperature. Two other classical signs of inflammation, swelling caused by increased vasopermeabilty and pain caused by released mediators on nerve endings, may also occur. Both UVB and UVA radiations elicit inflammation although the initiating photochemistry, cell signaling processes, and biochemical pathway processes involved are not identical. Erythema has been used as the endpoint for measuring the relative effects of different wavelengths in the UVB and UVA, usually expressed as an action spectrum.

Action Spectra

An action spectrum indicates which wavelengths most effectively produce a skin response. It is used to understand the basic science of a photobiologic as well as a therapeutic response because phototherapy is most efficient when the emission of the lamp matches the action spectrum for the beneficial response. An action spectrum is most accurately plotted as the reciprocal of the number of incident photons required to produce a given effect (Y-axis) against wavelength (X-axis). Conventionally in dermatology, the reciprocal of the minimum fluence (exposure dose) to produce the response is plotted versus wavelength. In an ideal case, the action spectrum corresponds to the absorption spectrum for the chromophore and can be used to identify the chromophore for a given photobiologic reaction. For example, an action spectrum for wheal formation in a patient with EPP showed a maximum at 400 nm, close to the maximum of protoporphyrin IX, which supports biochemical evidence for this porphyrin as the chromophore.18 Another example involves the action spectrum for the clearance of psoriasis, which was found to peak around 313 nm.3 This insight has led to the development of so-called “narrowband” UVB sources with peak output at 311 nm, and the use of the excimer 308-nm laser for the efficacious treatment of psoriasis. Absorption in this range suggests that DNA is a punitive chromophore, but given the different depths of penetration of the two wavelengths into skin and the varying clinical responses, it is possible the relevant chromophore for the therapeutic effect of UVB on psoriasis is not exclusively DNA. Moreover, in some cases, the action spectrum for a response may not directly correlate to the absorption spectrum of the responsible chromophore because the wavelengths actually reaching the chromophore are distorted by absorption and scattering of light by tissue above the level of the chromophore.

The action spectrum for erythema indicates that the most effective wavelengths present in sunlight are in the UVB range.1,19 UVA is roughly 1,000-fold less effective than UVB. The time in the sun required to produce sunburn is strongly influenced by many factors including skin pigmentation, season, geographical location, cloud cover and time of day. The relative contributions of UVB and UVA radiation from the sun to a sunburn response are not constant because the UVB to UVA ratio in the solar spectrum varies with the time of day, the season and other factors.

UVB-Induced Inflammation

Erythema is usually apparent 3–5 hours after UVB exposure, reaches a maximum intensity at 12–24 hours and generally resolves by 72 hours. However, when the UVB dose is high, erythema appears sooner, reaches a maximum at a later time and is more persistent, and the response is more likely to include pain or swelling. The minimal erythema dose or MED is the fluence that elicits minimal erythema at 24 hours. The UVB MED for fair skin (Fitzpatrick type I–II) using fluorescent UVB bulbs is about 30 mJ/cm2 (300 J/m2). Many biochemical and cellular events occur before the erythema is evident including increased blood flow, endothelial cell activation and elevated inflammatory mediator levels.20,21

Histology—UVB

The extent and time course of the histological changes observed after UVB varies with the magnitude of the exposure dose.22,23 Dyskeratotic cells, so-called “sunburn cells,” may appear in the epidermis as soon as 30 minutes after a 3 MED exposure. Sunburn cells are apoptotic keratinocytes that show condensed nuclear chromatin and eosinophilic cytoplasm. They increase in number with a maximum at about 24 hours and eventually (by 72 hours) form a band at the stratum corneum. Intercellular edema occurs early and may persist to 72 hours.

Dermal changes include early (30 minutes) and persistent (up to 72 hours) endothelial cell swelling, perivenular edema and degranulated mast cells.22 Neutrophils appear perivascularly shortly after UVB and the influx peaks at about 14 hours. A later mononuclear (macrophage) infiltrate persistes at least 48 hours.

Mechanism—UVB

Absorption of UVB photons by chromophores in skin leads to the production of inflammatory mediators and cytokines that characterize the sunburn response. The action spectrum for erythema closely correlates with the absorption spectrum of DNA suggesting that DNA is one of the important chromophores and that DNA photoproducts initiate at least partially the biochemical pathways leading to UV-induced inflammation.1,24 A role for DNA photoproducts is further supported by the observation that xeroderma pigmentosum patients, who have low ability to repair DNA, demonstrate prolonged UVB-induced erythema that can be reduced by treatment with a DNA repair enzyme.25 The biochemical pathways linking the DNA photoproducts with inflammatory mediators are still unclear. However, an action spectrum for TNF-α followed the DNA absorption spectrum suggesting formation of this proinflammatory cytokine is initiated by DNA photoproducts.26

In addition to DNA, other UVB-absorbing chromophores initiate formation of inflammatory mediators.27,28 A UVB-induced photoproduct of tryptophan initiates signaling events leading to expression of cyclooxygenase-2 (COX-2), which catalyzes formation of prostaglandins, which are associated with erythema.29 UVB exposure rapidly generates ROS that oxidize membrane lipids leading to proinflammatory cytokines.27,28,30 The role of ROS is supported by the ability of topical antioxidants to modulate UVB-induced erythema.12,31

Inflammatory Mediators—UVB

Prostaglandins and nitric oxide appear to be the major mediators for UVB-induced inflammation. Early studies showed that PGE2 levels in suction blister fluid increased before the appearance of erythema and continued to rise at least until 24 hours.22 More recently, numerous proinflammatory and anti-inflammatory eicosanoids were identified in suction blister fluid over a 72 hours period after UVB exposure.15 Prostaglandins with vasodilatation activity (PGE2, PGF2α, and PGE3) appeared at 24–48 hours, and the species with leukocyte chemoattractant activity [11-, 12-, and 8-monohydroxy-eicosatetraenoic acid (HETE)] were present to later times (4–72 hours). Interestingly, an anti-inflammatory species, 15-HETE, was maximally present at 72 hours. The first UVB-induced step in the formation of these eicosanoids is a rapid (<30 minutes) increase in the activity of phospholipase A2, which liberates arachidonic acid from cell membranes. Arachidonic acid is then acted on by cyclooxygenases and lipoxygenases to produce the eicosanoids. Inhibiting cyclooxygenases with indomethacin applied after UVB suppressed erythema.21,32

A role for nitric oxide in UVB-induced inflammation was established using NG-nitro-l-arginine methyl ester (L-NAME), a specific inhibitor of nitric oxide synthase, which suppressed erythema.21,33 Interestingly, L-NAME effectively suppressed erythema even when injected 48 hours after UVB indicating that nitric oxide was continuously synthesized. Histamine, however, which can trigger an inflammatory response, does not appear to be a major contributor to the UVB-induced sunburn34 even though it rises after UVB.22

Several proinflammatory cytokines, including TNF-α, IL-1, IL-6, and IL-8, are elevated after UVB exposure of human skin.35–38 The IL-1-like activity was detected by 1 hour after UVB36 and TNF-α was elevated by 4 hours, maximal at 15 hours, and then declined.35 Cytokines may play several roles in UVB-induced inflammation including recruiting leukocytes from the vasculature by acting as chemoattractants and as inducers of endothelial cell adhesion molecules that facilitate movement of cells across the capillary walls.

UVA-Induced Inflammation

Sunburn elicited by UVA has been much less thoroughly examined than UVB-induced erythema even though the amount of UVA in sunlight vastly exceeds (∼19-fold) the UVB portion. UVB and UVA contribute significantly to the effects of chronic sun (and tanning bed) exposure. Erythema is usually present at the end of the UVA exposure using clinical or laboratory light sources. This immediate response may fade or merge into a delayed erythema with a maximum between 6 and 15 hours depending on the fluence, irradiance and spectrum of the light source.20,39 The UVA MED in fair skin is generally 30–75 J/cm2.20,23,40

Histology—UVA

In contrast to UVB-induced erythema, the major histological changes after UVA exposure occur in the dermis. Most notably, the sunburn cells that characterize UVB-exposed epidermis are not apparent after UVA.23,39 Epidermal intercellular edema was present for at least 48 hours, and LCs decreased over 48 hours. In the dermis, neutrophils peaked at 3 hours, earlier than after UVB, and remained present for at least 48 hours. A lymphocytic infiltrate was also apparent throughout the dermis at all times after irradiation.39

Mechanisms—UVA

The differences in time course and histology between UVB and UVA-induced responses suggest that different mechanisms are involved. The erythema action spectrum supports this concept; the slope of the curve decreases at wavelengths longer than about 330 nm and a small shoulder is seen at about 360 nm.1,19 In addition, UVA-induced erythema was suppressed when the skin oxygen level was reduced whereas the UVB erythema was not affected.40 Although several UVA chromophores are present in skin (Fig. 90-6), the identity of the light absorber for UVA sunburn is still unidentified.

ROS play a significant role in UVA-induced inflammation, as suggested by the oxygen dependence of UVA erythema.40 Singlet oxygen and other ROS are detected by spectroscopy when skin is exposed to UVA.11,41 These reactive species are responsible for activating phospholipase A2, which releases arachidonic acid from membrane lipids and makes it available for enzymatic conversion into proinflammatory eicosanoids. ROS also initiate intracellular signal transduction pathways that activate the synthesis of proinflammatory cytokines. In addition, UVA has been shown to liberate nitric oxide very rapidly from storage forms in skin.42

Inflammatory Mediators—UVA

Exposure of skin to UVA increased the levels of proinflammatory prostaglandins as measured in suction blister fluid.43 A study in keratinocytes indicated that increased prostaglandin levels were caused by UVA-induced activation of PLA2.44 In vivo UVA stimulated an increase in the levels of PGD2, PGE2, and 6-keto-PGF1α at 5–9 hours, which returned to normal by 24 hours.43 Histamine also increased by 9–15 hours and returned to baseline by 24 hours. UVB, but not UVA, stimulated production of TNF-α from keratinocytes and in human skin.45,46

Vitamin D Photobiology

In 1928, Adolf Windaus garnered the Nobel prize in chemistry for his studies on the constitution of sterols and their connection with vitamins.47 The fat-soluble substance under his investigation was vitamin D; however, the story of vitamin D and rickets, its deficiency, actually dates back to antiquity based on careful study of writings and paintings. Conducting dietary studies with cod-liver oil on dogs in 1919, Mellanby first concluded that rickets stems from the absence of an “accessory food factor.”48 Three years later McCollum et al found that heated, oxidized cod-liver oil could cure rickets in rats.49 They named the new factor vitamin D because it was the fourth vitamin discovered.

Concurrently, an entirely different remedy for rickets appeared in the form of UV light. By the late nineteenth and early twentieth centuries, lack of fresh air and sunshine or lack of exercise was implicated in the etiology or rickets.50 In 1921, Hess and Unger observed that seasonal incidence of rickets seemed to parallel seasonal variations of sunlight.51 Chick independently observed that sunlight would cure rickets just as well as cod-liver oil.52 In 1919, Huldschinsky reasoned that artificial sunlight might well do the same as natural sunlight.53 Controlling for dietary intake and ambient UV exposure, he irradiated severely rachitic children with a UV-emitting quartz-mercury lamp and observed remarkable clinical and radiographic improvement, including fresh calcium deposition.

From cottonseed oil in 1925, Hess and his team isolated sitosterol, which was inactive against rickets in rats until it was irradiated by UV light.54 The discovery that irradiation of food, in particular of whole milk, could impart antirachitic potency led to tremendous advances in public health, producing a rapid decline in the prevalence of rickets in children. With amazing foresight, Hess hypothesized that cholesterol in skin is activated by UV-irradiation and rendered antirachitic. The complete photochemical and thermal reaction steps in the vitamin D pathway were finally elucidated in 1955 by Velluz.55 The exact sequence of steps leading to the cutaneous photoproduction of cholecalciferol was reported in a comprehensive paper by Holick in 1980.56

Functions of Vitamin D

Vitamin D regulates calcium and phosphorus metabolism.56 Its major role is to increase the flow of calcium into the bloodstream, by promoting absorption of calcium and phosphorus from the intestines, and reabsorption of calcium in the kidneys, thus enabling normal mineralization of bone and muscle function. It affects serum alkaline phosphatase level. Vitamin D also inhibits the proliferation of T-cells and the maturation of dendritic cells (DCs) along with affecting keratinocyte function.

Vitamin D deficiency results in impaired bone mineralization that leads to bone softening diseases, rickets in children and osteomalacia in adults, and possibly contributes to osteoporosis.57 Deficiency can arise from inadequate intake coupled with inadequate sunlight exposure, disorders that limit its absorption, or conditions that impair conversion of vitamin D into active metabolites, such as liver or kidney diseases. Most vulnerable to low vitamin D levels are elders, individuals living at high latitudes with long winters, obese persons, and all individuals with dark skin pigmentation living at high latitudes.58–65

Toxicity from excess vitamin D may manifest itself as hypercalciuria or hypercalcemia, the latter causing muscle weakness, apathy, headache, confusion, anorexia, irritability, nausea, vomiting, and bone pain, and potentially leading to complications such as kidney stones and kidney failure. Chronic toxicity effects include the aforementioned symptoms along with constipation, anorexia, abdominal cramps, polydipsia, polyuria, backache, and hyperlipidemia. Findings may also include calcinosis, followed by hypertension and cardiac arrhythmias (due to shortened refractory period). Although information about the effects of high doses of vitamin D is limited, 10,000 IU daily is a safe upper limit for adult intake.66 Chronic toxic dose is more than 50,000 IU/day in adults.

There are two major sources of vitamin D, one is diet and the other is skin. When obtained exogenously from food or supplements, vitamin D is absorbed in the small intestine. Natural food sources rich in vitamin D include certain oily fish such as salmon, mackerel, tuna, herring, catfish, cod, sardines and eel, butter, margarine, yogurt, liver, liver oil, and egg yolks, but at least in the United States, most dietary vitamin D comes from the fortification of food like cereal, milk, and orange juice.67 Fortified milk, for example, typically provides 100 IU per 8 oz glass, or only a quarter of the estimated adequate daily intake for adults. To get the recommended daily allowance, most Americans have to take supplemental vitamin D either alone, or with calcium or in a multivitamin.

Biochemical Pathway

(Fig. 90-8.)

As a result of UVB exposure on the skin, provitamin D3 (7-dehydrocholesterol, a precursor of cholesterol) is rapidly converted to previtamin D3, which spontaneously isomerizes to vitamin D3, entering the circulation on a binding protein and joining dietary D2 (or ergocalciferol) and D3 (cholecalciferol) absorbed from the gut.56 Once reaching the liver, these undergo passive hydroxylation in the endoplasmic reticulum of the hepatocytes, a process that requires NADPH, O2 and Mg2+. The product 25-hydroxyvitamin D3 [25(OH)D3 (calcidiol)] is stored in the hepatocytes until it is needed when it can be released into the plasma to make its way to the proximal tubules of the kidneys, where it is acted upon by 25(OH)D-1-α-hydroxylase, an enzyme whose activity is increased by parathyroid hormone and low PO42−. People with kidney disease may not be able to convert vitamin D to its active form. Following this conversion, 1,25-hydroxyvitamin D3 [1,25(OH)2D3 (calcitriol)] is released into the circulation, and by binding to a carrier protein in the plasma, vitamin D-binding protein (VDBP), it is transported to various target organs.

Action Spectrum for Vitamin D Formation in Skin

Action spectra studies show that wavelengths the most effective at bringing about the photosynthesis of cutaneous vitamin D lie in the range of 295 nm to 315 nm, which ironically are some of the same wavelengths most responsible for photocarcinogenesis.68 Optimal synthesis occurs in a very narrow band of UVB spectra between 295–300 nm with peak isomerization at 297 nm.68,69 With a UV Index of at least 3, which occurs daily in the tropics and almost never at high latitudes, adequate amounts of vitamin D3 can be made in the skin after 10–15 minutes of sun exposure at least two times per week to the face, arms, hands, or back without sunscreen. In Boston, sunlight exposure from November through February is inadequate to result in significant cutaneous vitamin D production.65 The available flux of UVB for the synthesis of vitamin D depends on all the factors that determine the UV Index, including time of day, cloud cover, smog, shade, reflection from nearby water, sand or snow, latitude, altitude, and season of the year. Of course, individual factors such as age (vitamin D production decreases when older70), body mass index, clothing worn, amount of skin exposed also have an effect.71 Individuals with higher skin melanin content require more time in sunlight to produce the same amount of vitamin D as individuals with lower melanin content.

According to Holick, human beings make at least 10,000–25,000 units of vitamin D upon full body exposure to the sun at one minimal erythemal dose.72,73 Vitamin D production in the skin occurs within minutes and is already maximized before your skin turns pink. Exposure to sunlight for extended periods of time does not normally cause vitamin D toxicity. Within 20 minutes of sun exposure in fair-skinned individuals (1–3 hours for pigmented skin) the concentration of vitamin D precursors produced in the skin reaches an equilibrium, and excess vitamin D simply degrades as fast as it is generated.

Optimal Levels of Vitamin D

Ninety years after its discovery, vitamin D has become a subject of intense interest to dermatologists in part due to the fact that its deficiency is being increasingly recognized and that its deficiency may well have important health consequences beyond those related to bone health. As dermatologists counsel patients to employ sunscreens in order to minimize subsequent risk of skin cancer, they should be aware that sun protection recommendations may impede the efficacy of vitamin D photosynthesis in the skin.74 Several studies have all shown that most people, even those living in year-round sunny locations such as Florida and Australia have low vitamin D levels.75–79 In addition, other studies have demonstrated that minimal or even moderate UVB exposures may not be sufficient to maintain 25-hydroxy vitamin D [25(OH)D] levels at the now widely quoted “sufficient level” of 30 ng/mL.75,80 Consequently, relying solely on ambient exposure appears not to be sufficient as a means to obtain adequate vitamin D. Nearly all randomized trials that have proven the beneficial effects of increased vitamin D have achieved that increase by oral supplementation. It seems most likely that skin cancer prevention recommendations in the form of sunscreen use and judicious use of protective clothing did not cause the problem of low vitamin D levels and equally unlikely that urging unprotected ambient exposure will not fully solve this year-round problem either unless dietary vitamin D supplementation is recommended.

Low levels of vitamin D are clearly associated with falls and fractures in the elderly.81–86 As to whether vitamin D supplementation can reduce these complications, there is controversy with reports showing a benefit while others do not. Capturing more attention as of late is the mounting evidence that vitamin D plays a role in preventing colorectal cancer and to a lesser extent cancers of the breast and prostate.87–92 Other malignancies have been associated with vitamin D insufficiency in observational studies. Moreover, multiple other health problems have been shown to be inversely correlated in frequency with vitamin D levels in some studies. Examples include diabetes mellitus,93,94 multiple sclerosis,95 hypertension,96 myocardial infarction,97 cardiovascular,98 seasonal affective disorder,99 chronic pain,100 peripheral artery disease,101 cognitive impairment, which includes memory loss,102 and infections,103,104 among others. There is an association between low vitamin D levels and Parkinson's disease.105 Current evidence is simply insufficient to infer a causal relationship in any of these conditions.

What is an optimal level of vitamin D? Risk of cardiovascular events in the Framingham Offspring cohort study suggested levels as low as about 20–25 ng/mL.106 The national representative National Health and Nutrition Examination Survey data suggested that overall mortality was minimized at levels of about 30–40 ng/mL.101 Given concerns about potential confounding factors, we cannot be certain about these estimates of optimal levels, but the preponderance of published evidence would support values of 25–40 ng/mL.107 Thirty ng/mL (75 nmol/L) is a common estimate given current knowledge.108

Clinicians can measure individual levels and recommend supplementation accordingly. Common guidelines include 400–1,000 IU daily with or without calcium or in a multivitamin, 10,000 IU weekly or every 10 days, or 50,000 IU every month. After correction of their vitamin D status with oral vitamin D, patients should have a repeat test of their 25(OH)D level to confirm that they are in the normal range. Such measurement is necessary since there is great variation in the increment of 25(OH)D achieved from a given dose of vitamin D supplement. If a patient remains persistently low, despite several attempts at correction with oral vitamin D, there may be malabsorption and referral to a gastroenterologist is in order. A trial of UVB light therapy may be considered to improve vitamin D status. For patients with 25(OH)D levels less than 15 ng/mL, one usually prescribes oral supplements of 50,000 IU vitamin D weekly for 8 weeks and then switches to standard maintenance doses. Efficacy of vitamin D supplementation is dependent on body mass index (BMI).57 Overweight and obese patients with hypovitaminosis D require higher doses of vitamin D to achieve vitamin D repletion compared to individuals with normal body weight.

In short, we can conclude that vitamin D supplementation appears to enhance musculoskeletal health and may possibly reduce mortality in some groups. We have evidence that vitamin D may have an influence on cancer, cardiovascular disease, autoimmune disease, and infection, but this evidence, although suggestive, is quite variable in strength and does not allow the conclusion that any of these associations are proven. These are subjects of ongoing investigation.

Skin cancer is caused largely by direct damage to DNA, leading to specific gene mutations. However, for the development of clinically apparent skin cancer, a second UVB-induced mechanism is required, namely, UVB-induced immunosuppression.109,110 The connection between sun exposure and skin cancer was realized first in the beginning of the last century when physicians observed skin tumors that grew preferentially at sites that had been extensively sun exposed. This was followed in the 1970s by a seminal study by Kripke,111 who discovered a connection between immunosuppression and photocarcinogenesis. In a series of experiments, she demonstrated that UV-induced tumors, which had been transplanted onto syngeneic recipient mice, were rejected if the recipient mice had not been exposed to UVB. This result implies that UV-induced tumors have a highly antigenic phenotype111 and can be rejected by a functioning immune system. Treatment with known immunosuppressive drugs, or—this was the novel finding—with low doses of UVB, led to failure of tumor rejection.112,113 Both aspects (i.e., the strong antigenicity of UV-induced tumors and the immunosuppression by UVB) have triggered vigorous research activities in the field now termed photoimmunology. Numerous studies tried to clarify the cellular and molecular mechanisms.114 These have turned out to be highly complex (see Fig. 90-9). Results sometimes appear contradictory, likely due to use of various experimental approaches and model systems.

Model Systems for Ultraviolet B-Induced Immunosuppression

The standard model for studying UVB-induced immunosuppression is the delayed-type hypersensitivity (DTH) to antigens and the contact hypersensitivity reaction (CHS). The ability of UVB irradiation to block either induction or elicitation of allergic sensitization can then be quantified; typically a hapten such as dinitrofluorobenzene or oxazolone is applied to the skin as the stimulus and intensity of the subsequent immune response is measured as tissue swelling. DTH responses have become the model of choice to study the immunosuppressive effects of UVB radiation, as the tumor-associated antigens of UVB-induced tumors are only poorly characterized and testing DTH responses is less time-consuming and less laborious than examining photocarcinogenesis. UVB radiation has also been shown to affect immune responses involved in the pathogenesis of viral, parasitic, fungal, and bacterial infections. UVB-mediated immunosuppression is of two types: (1) local immunosuppression in which the immune response to antigens applied at the irradiated site is abrogated, and (2) systemic immunosuppression in which the immune response to antigens applied to unexposed sites is impaired (Fig. 90-10).115 However, some mice are more susceptible to UVB-induced immunosuppression than others, with the genetic loci for lipopolysaccharide, Lps, and tumor necrosis factor, Tnf, strongly influencing the response.116

Molecular Targets in Photoimmunosuppression

DNA is a chromophore for UVB radiation and can be a direct target (see Fig. 90-6). UVB radiation generates a variety of photoproducts, most commonly CPDs6–4 and photoproducts. Pyrimidine dimer formation is a direct cause of UV-induced immunosuppression (Fig. 90-9). The cells of the opossum Monodelphis domestica contain endogenous photoreactivating enzymes, which are capable of repairing CPDs, and UVB does not induce immunosuppression in these animals on exposure to photoreactivating light immediately after irradiation.117 Also in mice, a correlation between CPDs and UVB-induced systemic immunosuppression exists.118

UVB irradiation leads to CPDs in antigen-presenting cells (APCs) and impairment of their antigen-presenting capacity. The damage persists for several days, and the damaged cells migrate from the skin to lymph nodes. Treatment at the site of UVB irradiation with liposomes containing the DNA repair enzyme T4 endonuclease V prevents the impairment in antigen presentation.119 The role of DNA damage in UVB-induced immunosuppression was also confirmed in human skin, in which topical application of photolyase-containing liposomes to UVB-exposed sites prevents UVB-induced immunosuppression.120

In addition to direct DNA damage, the generation of ROS is likely to be involved in photoimmunosuppression (see Fig. 90-9). Accordingly, topical application of a polyphenolic fraction isolated from green tea before and after UVB irradiation to mouse skin protects against UV-induced immunosuppression,121 an effect most likely mediated by the antioxidant epigallocatechin-3-gallate, the component of extract that has the ability to reduce oxidative stress, inflammatory responses, and skin cancer. It should be noted that such secondary plant metabolites affect cell signaling, however, so that protection against UVB-induced events may not exclusively be explained by antioxidant effects.

The cell membrane of immunocompetent skin cells is another target for UV radiation. An important function of the cell membrane is to transfer signals from outside into the cell via membrane receptors. UVB exposure leads to clustering and internalization of cell surface receptors for epidermal growth factor, TNF, and IL-1 in the absence of the respective ligands. Clustering results in strong activation of stress-induced c-Jun NH2-terminal kinases, which are members of the mitogen-activated protein kinase family122 (see Fig. 90-9). Conceivably, such aberrant receptor clustering may subvert signaling pathways normally used by growth factors and cytokines, eventually contributing to the immunological response. The chromophore for these UVB-induced membrane alterations is cytoplasmic tryptophan. The resulting tryptophan photoproduct(s) bind to the transcription factor arylhydrocarbon receptor (AhR), which in its inactive form is part of a cytoplasmic complex including c-src and the heat shock protein Hsp90. Upon binding of the tryptophan photoproduct to the AhR, the complex dissociates and the AhR translocates into the nucleus whereas c-src translocates to the cell membrane where it subsequently causes the previously mentioned cell membrane/surface receptor activation.29

UVB radiation can lead to lipid peroxidation, including cell membrane lipids. Phosphatidylcholine, an important membrane lipid, can be oxidized to platelet-activating factor(PAF)-like lipids that bind to PAF receptors and activate cytokine synthesis. A role of UVB-induced PAF formation in photoimmunosuppression is suggested by the observation that comparable immunosuppression can be induced in mice when they are irradiated with UVB, injected with carbamyl-PAF, or injected with phosphatidylcholine that had been exposed to UVB under atmospheric oxygen. The immunosuppressive properties of all three treatments were equally blocked by specific PAF receptor antagonists, indicating that UVB-induced PAF-like lipids can trigger the PAF receptor in a way similar to PAF. In turn, PAF receptor activation stimulates a variety of downstream effects, including the activation of mitogen-activated protein kinase pathway and synthesis of immunosuppressive cytokines (see Fig. 90-9).123 Interestingly, UVB radiation-induced immunosuppression as well as formation of PAF were enhanced in mice, which were unable to take up the osmolyte taurine. Thus, similar to nucelotide excision repair, taurine uptake is critically involved in protection of the skin against UVB radiation-induced immunosuppression.124

An extracellular chromophore that mediates UVB-induced immunosuppression is UCA (see Fig. 90-6).125 UV radiation isomerizes trans-UCA to cis-UCA dose-dependently until a balanced photostationary state is reached. A number of independent studies confirmed the importance for cis-UCA in UVB-induced immunosuppression, although the exact mechanism of action is not clear at present.

Cellular Events Involved in Photoimmunosuppression

UVB radiation of keratinocytes alters their expression of surface molecules and induces the synthesis and secretion of a whole set of immunomodulatory soluble factors, including IL-1, IL-6, IL-8, TNF-α, and PGE2 (see Fig. 90-9).126 Evidence for keratinocyte-derived soluble factors with immunosuppressive function came from experiments showing that either serum factors from UVB-exposed mice or supernatants from UVB-irradiated murine epidermal cells suppressed hypersensitivity reactions in mice. Subsequently, it was found that TNF-α and IL-10, two potent immunosuppressive cytokines, are produced by UVB-irradiated keratinocytes as well.125,128 IL-10 is also released by CD11b+ macrophages in human skin.129 Injection of TNF-α mimics UVB-induced alterations on Langerhans cells (LCs). The administration of a neutralizing antibody to TNF-α before irradiation abrogates partially UVB-induced accumulation of DCs in draining lymph nodes (DLNs) and suppression of contact hypersensitivity. The release of TNF-α is of special importance for local immunosuppression, and, as mentioned above, the Tnfa gene locus contributes to UVB susceptibility.

IL-10 is a T-helper cell type 2 (Th2) cytokine and abrogates the production of Th1 cytokines, in particular, interferon-γ (IFN-γ). UVB-induced IL-10 production by keratinocytes in murine or macrophages in human skin130 shifts the immune response from a Th1 to a Th2 type. This can explain why Th1-mediated cellular immune reactions are impaired by UV radiation.

The interactions between the different cytokines are not fully understood, but IL-10 certainly is a major player. For instance, anti-IL-10 antibodies neutralize the suppressive effects of supernatants from UVB-irradiated keratinocytes on DTH in mice.127 The main mechanism of IL-10 is probably inhibiting the antigen-presenting capacity of LCs. In parallel, UV-exposed keratinocytes release PAF and PAF-like lipids, which trigger the PAF receptor, as described in Section “Molecular Targets in Photoimmunosuppression.”29 Presumably, PAF and its receptor enhance the production of COX-2, PGE2, IL-4, and IL-10. IL-10 production can be reversed by IL-12. Incidentally, IL-12 can restore UV-impaired CHS.130 Injection of mice with IL-12 after UVB exposure prevented UVB-induced immunosuppression and suppressed partially UVB-induced tolerance obtained by resensitization 14 days after the UV irradiation. This IL-12 effect is most likely linked to the capacity of this cytokine to enhance the repair of UVB radiation-induced DNA photoproducts, which form the molecular basis for IL-10 production by irradiated keratinocytes.

Cellular Mediators of Photoimmunosuppression: Langerhans Cells, Macrophages, and T Cells

LCs, 2%–5% of epidermal cells, are a DC subset that originates from bone marrow precursors and populates the epidermis. They are the “professional” skin-specific APCs and ingest antigen locally in the skin, but lack costimulatory capacity. After antigen uptake, LCs migrate to the DLN, where they mature to potent stimulators for antigen-specific T cells, expressing high levels of major histocompatibility complex molecules and costimulatory molecules B7.1 and B7.2 and intercellular adhesion molecule-1 on their cell surface (see Chapter 10).

In the 1980s, it was demonstrated that epidermal LCs play a pivotal role in the induction and regulation of CHS reactions. LC density is an important factor for the induction of CHS responses, and UV radiation leads to the disappearance of LC from irradiated sites.131 In these studies, mice were sensitized with the hapten dinitrofluorobenzene either on skin with typical amounts of LC or on LC-deficient skin. A significantly impaired CHS response was observed when the mice were sensitized on LC-deficient skin. The capacity to deplete LCs temporarily from the skin appears to be an important feature of UVB irradiation. Moreover, UV exposure results in reduced surface expression of costimulatory molecules Ia, B7.1, B7.2, and intercellular adhesion molecule-1 (see Fig. 90-9).132,133 The depletion of these molecules might account for or contribute to immunosuppression. The loss of LCs in the skin is accompanied by an increase in DCs in the DLN. 134 Fluorescein isothiocyanate+ Ia+ CPD+ cells in the DLNs were found after painting the skin with fluorescein isothiocyanate and irradiating subsequently with UV. This demonstrates that LCs indeed leave the skin after UV exposure. UVB possibly induces LC migration in the absence of exogenous antigen, although the migratory effect is enhanced by sensitization with antigen. The ultimate fate of chronically irradiated LCs remains to be examined. UVB-induced LC migration was also observed in human skin in vivo.135

In addition to reducing the number of LCs in the skin, UVB radiation affects LC function by impairing the antigen-presenting capacity in DTH and CHS. UVB-irradiated LCs activate preferentially CD4+ Th2 cells.136 In contrast, they do not activate CD4+ T cells of the Th1 subset, but induce clonal anergy.137 Another cellular player in UV-induced immunosuppression is CD11b+ macrophages, which infiltrate UVB-irradiated human skin (see Fig. 90-9). These cells produce IL-10, and their deletion by antibodies makes UV-irradiated mice permissive again for CHS.138

In addition to immunologic unresponsiveness to haptens applied after irradiation, UVB exposure-induced hapten-specific tolerance (i.e., resensitization with the same hapten at later time points) likewise fails.139 In adoptive transfer experiments, it was shown that UVB induces T-suppressor cells. Elmets et al140 prepared single cell suspension of the spleen and lymph nodes from UVB-exposed and hapten-treated mice, injecting the cell suspension into naive mice. When the recipient mice were sensitized and challenged with the same hapten that was used for the donor mice, they showed significantly impaired CHS responses.

The phenotype of the regulatory T cells and their mechanism of action are not well understood (Fig. 90-11). Results obtained so far are difficult to compare. Different experimental setups yielded partly contradictory results. There is evidence to suggest that CD8+ T cells are important mediators of UVB-induced immunosuppression in the local immunosuppression.141 In addition, T cells (not defined by CD4 or CD8) that express cytotoxic T lymphocyte antigen (CTLA)-4 on their surface transfer UV-induced tolerance.142 CTLA-4 appeared to be of functional relevance, because in vivo injection of anti-CTLA-4 antibodies blocked transfer of suppression. The physiologic function of CTLA-4 is to end immune responses by shutting down activated T cells, so CTLA-4 is an important molecule in immune regulation. On stimulation with antigen-bearing DCs, CTLA-4+ T cells secrete high levels of IL-10, transforming growth factor-β, and IFN-γ, but low levels of IL-2 and no IL-4. Other studies showed that CD4+ T cells from UV-exposed mice are capable to suppress CHS when injected into untreated recipient mice.143 These CD4+ T cells produce IL-10, but no IL-4 and IFN-γ on APC coculture and inhibit IL-12 production by APCs.

Apart from CD4+ regulatory T cells, natural killer T cells with regulatory capacities were found in UVB-induced systemic immunosuppression (see Fig. 90-10). In particular, CD3+, CD4+, and DX5+ CD1-restricted natural killer T cells have been described, which secrete IL-4 and can adoptively transfer suppression of tumor rejection as well as DTH responses.134

Ultraviolet Radiation Effects on Innate Immunity

In the skin, in addition to slow but specific adaptive immune responses, the more primitive innate immune response can be generated as well. Essential components of the innate immune response are neutrophils, eosinophils, natural killer cells, mast cells, cytokines, complement, and, in particular, amtimicrobial peptides and proteins. It is important to note that the immunosuppressive properties of UV radiation refer exclusively to the adaptive immune response. In contrast, UV radiation has been shown to induce rather than suppress the innate immune response by increasing the production of the antimicrobial peptides human β-defensin (HBD)-2, -3, ribonuclease 7 and psoriasin in human keratinocytes and skin in vitro and in vivo.144 This may explain why T-cell-mediated immune reactions are suppressed on UV exposure but not host defense reactions against bacterial attacks.

Photoimmunology: Concluding Remarks

UV radiation is a permanent environmental threat, and adaptive responses against its damaging effects have been developed in all living organisms. From a theoretic point of view, it is therefore tempting to speculate that photoimmunosuppression serves a physiologic role as an adaptive response of the skin to UV-induced modifications of proteins and an immunologic response to so-called neoantigens and, as a consequence, chronic inflammation. Under these conditions, the higher risk of skin cancer that appears to accompany UV-induced immunosuppression appears of less significance in evolutionary terms, as skin cancers occur principally in late life, after the reproductive period. From a practical point of view, however, photocarcinogenesis poses a serious health problem. Accordingly, modern approaches to prevent UV-induced immunosuppression include broad-spectrum UV filters, liposomally encapsulated DNA repair enzymes, antioxidants, and osmolytes.145